Publication date: 10th April 2024
All-solid-state batteries relying on Li metal as negative electrode material and a ceramic electrolyte might severely suffer from unwanted interfacial processes between the active materials and the ionic conductor. The dynamic properties of the ionic-electronic, that is, mixed conducting interphase (MCI) [1] formed are mostly unknown. Here, we used Li1.3Al0.3Ti1.7(PO4)3 (LATP) as a model electrolyte that is known to form an MCI when in contact with metallic Li or any other Li source. We prepared Li1.3+xAl0.3Ti1.7(PO4)3 with x = 0.2, 0.6 and 1.3 via ex situ chemical lithiation to mimic the formation of MCIs taking place otherwise operando. The preparation of large amounts of lithiated LATP with controlled Li contents allowed us to use nuclear and electric techniques to study local structures and ionic/electronic dynamics in detail. X-ray powder diffraction (XRPD), X-ray photoelectron spectroscopy (XPS), and 7Li, 27Al magic angle spinning nuclear magnetic resonance (NMR) point to the formation of a core-shell two-phase morphology with the Li-rich paramagnetic Li3Al0.3Ti1.7(PO4)3 phase covering the unlithiated Li-poor regions. Importantly, chronoamperometric polarization revealed an increase of the originally poor electronic conductivity σeon of 6.5 × 10−12 S cm−1 (293 K) by ca. 3 orders of magnitude, hence reaching the order of 6.6 × 10−9 S cm−1 if x = 0.6. At the same time, we observe a slight increase in bulk ion conductivity, which is also probed by 7Li NMR spin-lattice realxation measurements [2]. At even higher loadings (x = 1.3), a decrease in electronic (and ionic) conductivity is seen for Li3Al0.3Ti1.7(PO4)3, i.e., not exceeding alarming values for σeon. Our findings help understand the important interfacial side reactions in all-solid-state batteries. A detailed characterization of the two-phase system, including the separation of electronic and ionic transport processes, is essential to assess the severity of damage through MCI formation and to discuss whether any strategies to mitigate them are necessary at all.
We thank the FFG (safeLIB) and the DFG (WI 3600, NMR spectrometer) for financial support.